Robust TV news story identification via visual characteristics of anchorperson scenes

Chia Hung Yeh, Min Kuan Chang, Ko Yen Lu, Maverick Shih*

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper, a new scheme for TV news segmentation via exploring the efficient visual features is proposed especially for TV news which contains lots of changeful background of anchorperson shots. The proposed scheme can be divided into two parts: probable anchorperson shot detection and real anchorperson detection. Our proposed method can efficiently detect anchorperson shots even though anchorperson shots contain changeful background and anchorperson position variation. Meanwhile, non-anchorperson shots can be robustly excluded from report shots such as interview scenes. Experimental results are given to demonstrate the feasibility and efficiency of the proposed techniques.

Original languageEnglish
Title of host publicationAdvances in Image and Video Technology - First Pacific Rim Symposium, PSIVT 2006, Proceedings
PublisherSpringer Verlag
Pages621-630
Number of pages10
ISBN (Print)354068297X, 9783540682974
DOIs
Publication statusPublished - 2006
Externally publishedYes
Event1st Pacific Rim Symposium on Image and Video Technology, PSIVT 2006 - Hsinchu, Taiwan
Duration: 2006 Dec 102006 Dec 13

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4319 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other1st Pacific Rim Symposium on Image and Video Technology, PSIVT 2006
Country/TerritoryTaiwan
CityHsinchu
Period2006/12/102006/12/13

Keywords

  • Anchorperson detection
  • Face recognition
  • Skin color
  • TV news segmentation
  • Visual features

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint

Dive into the research topics of 'Robust TV news story identification via visual characteristics of anchorperson scenes'. Together they form a unique fingerprint.

Cite this